122 



Puccinia Gramineae. 



other authors, and a double asterisk indicates those which 

 by Eriksson in Europe to produce aecidia on the Barberry 

 were sown upon it : 



Agropyron divergens Nees 

 *A. scabrum Beauv. 



Alopecurus geniculatus L. 

 Amphibromus neesii Stead. 

 *Avena t'atua L. 

 **A. saliva L. 



Beckraannia erucaeformis Host. 

 *Briza minor L. 



Bromus racemosus L. . . . 

 **B. secalinus L. 



B. sterilis L. ... 

 **Dactylis glomerata L. 



Deyeuxia quadriseta Benth. 



Echinopogon ovatus Beauv. 



Elymus striatus Willd. 



E. virginicus L. 



Festuca broraoides L. 



Glyceria dives F.v.M. ... 



G. stricta Hook.f. 

 **Hordeum murinum L. 



H. secalinum Schreb. ... 

 **H. vulgare L. 

 **Phalaris canariensis L. 



P. minor Retz. 

 **Secale cereale L. 



Triticum polonicum L. 

 **T. vulgare Vill. 



have been found 

 when their spores 



II., III. 

 II., III. 

 II. 



II., III. 

 II., III. 

 II., III. 

 II., III. 

 II. 

 II. 

 II. 



II., III. 

 II., III. 

 II., III. 

 II., III. 

 II. 



II.. III. 

 II., III. 

 II., III. 

 II. 



II., III. 

 II., III. 

 II., III. 

 II., III. 

 II., III. 

 II., III. 

 II., III. 

 II., III. 



While P. graminis has been found on the above grasses, it must not be 

 ^assumed that the one will affect the other if growing alongside of each other. 

 It used to be considered that the same species would attack indiscriminately 

 any of its hosts, but it has now been proved that a selection or specialisation 

 goes on, and that the same morphological species living on different hosts is 

 not identical in the sense of affecting all alike. By means of infection ex- 

 periments with uredospores and aecidiospores obtained from definite teleuto- 

 spores, Eriksson arrived at this result, that P. graminis resolved itself into 

 a series of specialised forms, each of which was confined to a definite and 

 circumscribed series of host-plants. 



As the result of experiments so far, six biologically distinct forms are 

 constituted : 



1. Secalis On Rye. 



2. Avenae On Oat. 



3. Tritici On Wheat. 



4. Airae On Aira. 



5. Agrostidis On Agrostis. 



6. Poae On Poa. 



When time and opportunity offer, I hope to carry out experiments to 

 determine how far the rusts of the P. graminis type on the various grasses 

 are capable of infecting each other. 



(Plates I., Figs. 2, 5, 7, 8 ; XIV., Figs. 113-122 ; XL., Fig. 301 ; 

 XLIIL, Fig.' 311 ; Plate A., Figs. 3, 4.) 



